{"id":"https://openalex.org/W7106166467","doi":"https://doi.org/10.1109/iccad66269.2025.11240633","title":"COBRA: Algorithm-Architecture Co-optimized Binary Transformer Accelerator for Edge Inference","display_name":"COBRA: Algorithm-Architecture Co-optimized Binary Transformer Accelerator for Edge Inference","publication_year":2025,"publication_date":"2025-10-26","ids":{"openalex":"https://openalex.org/W7106166467","doi":"https://doi.org/10.1109/iccad66269.2025.11240633"},"language":null,"primary_location":{"id":"doi:10.1109/iccad66269.2025.11240633","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccad66269.2025.11240633","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/ACM International Conference On Computer Aided Design (ICCAD)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Ye Qiao","orcid":null},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ye Qiao","raw_affiliation_strings":["University of California,Department of Electrical Engineering and Computer Science,Irvine,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Department of Electrical Engineering and Computer Science,Irvine,USA","institution_ids":["https://openalex.org/I204250578"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhiheng Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhiheng Chen","raw_affiliation_strings":["University of California,Department of Electrical Engineering and Computer Science,Irvine,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Department of Electrical Engineering and Computer Science,Irvine,USA","institution_ids":["https://openalex.org/I204250578"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yian Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yian Wang","raw_affiliation_strings":["University of California,Department of Electrical Engineering and Computer Science,Irvine,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Department of Electrical Engineering and Computer Science,Irvine,USA","institution_ids":["https://openalex.org/I204250578"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yifan Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yifan Zhang","raw_affiliation_strings":["University of California,Department of Electrical Engineering and Computer Science,Irvine,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Department of Electrical Engineering and Computer Science,Irvine,USA","institution_ids":["https://openalex.org/I204250578"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yunzhe Deng","orcid":null},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yunzhe Deng","raw_affiliation_strings":["University of California,Department of Electrical Engineering and Computer Science,Irvine,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Department of Electrical Engineering and Computer Science,Irvine,USA","institution_ids":["https://openalex.org/I204250578"]}]},{"author_position":"last","author":{"id":null,"display_name":"Sitao Huang","orcid":null},"institutions":[{"id":"https://openalex.org/I204250578","display_name":"University of California, Irvine","ror":"https://ror.org/04gyf1771","country_code":"US","type":"education","lineage":["https://openalex.org/I204250578"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sitao Huang","raw_affiliation_strings":["University of California,Department of Electrical Engineering and Computer Science,Irvine,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Department of Electrical Engineering and Computer Science,Irvine,USA","institution_ids":["https://openalex.org/I204250578"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I204250578"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"8"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.5853999853134155,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.5853999853134155,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10502","display_name":"Advanced Memory and Neural Computing","score":0.07240000367164612,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.04650000110268593,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/binary-number","display_name":"Binary number","score":0.7348999977111816},{"id":"https://openalex.org/keywords/software-deployment","display_name":"Software deployment","score":0.5396000146865845},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.45320001244544983},{"id":"https://openalex.org/keywords/edge-device","display_name":"Edge device","score":0.44920000433921814},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.3847000002861023},{"id":"https://openalex.org/keywords/binary-independence-model","display_name":"Binary Independence Model","score":0.3790999948978424}],"concepts":[{"id":"https://openalex.org/C48372109","wikidata":"https://www.wikidata.org/wiki/Q3913","display_name":"Binary number","level":2,"score":0.7348999977111816},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6912999749183655},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.5396000146865845},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.45320001244544983},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.44920000433921814},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.4223000109195709},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.3847000002861023},{"id":"https://openalex.org/C37061001","wikidata":"https://www.wikidata.org/wiki/Q3531721","display_name":"Binary Independence Model","level":3,"score":0.3790999948978424},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.37610000371932983},{"id":"https://openalex.org/C64452783","wikidata":"https://www.wikidata.org/wiki/Q1524945","display_name":"Ternary operation","level":2,"score":0.37450000643730164},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3562000095844269},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3547999858856201},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.3273000121116638},{"id":"https://openalex.org/C13164978","wikidata":"https://www.wikidata.org/wiki/Q600158","display_name":"Hardware acceleration","level":3,"score":0.32170000672340393},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3140999972820282},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.31049999594688416},{"id":"https://openalex.org/C24326235","wikidata":"https://www.wikidata.org/wiki/Q126095","display_name":"Electronic engineering","level":1,"score":0.3075999915599823},{"id":"https://openalex.org/C2778456923","wikidata":"https://www.wikidata.org/wiki/Q5337692","display_name":"Edge computing","level":3,"score":0.26089999079704285},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.2596000134944916},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.2535000145435333}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/iccad66269.2025.11240633","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccad66269.2025.11240633","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/ACM International Conference On Computer Aided Design (ICCAD)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7","score":0.895548403263092}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":16,"referenced_works":["https://openalex.org/W2585560244","https://openalex.org/W2963145956","https://openalex.org/W2964242827","https://openalex.org/W3098576111","https://openalex.org/W3130641740","https://openalex.org/W3173374050","https://openalex.org/W3177265267","https://openalex.org/W3184454880","https://openalex.org/W4313564414","https://openalex.org/W4317795323","https://openalex.org/W4383108457","https://openalex.org/W4386763770","https://openalex.org/W4393178455","https://openalex.org/W4400231005","https://openalex.org/W4406769276","https://openalex.org/W4409282453"],"related_works":[],"abstract_inverted_index":{"Transformer-based":[0],"models":[1],"have":[2],"demonstrated":[3],"superior":[4],"performance":[5],"in":[6,24,112],"various":[7],"fields,":[8],"including":[9],"natural":[10],"language":[11],"processing":[12],"and":[13,21,27,53,102,123,139],"computer":[14],"vision.":[15],"However,":[16,56],"their":[17,30],"enormous":[18],"model":[19],"size":[20],"high":[22],"demands":[23],"computation,":[25],"memory,":[26],"communication":[28],"limit":[29],"deployment":[31,48],"to":[32,66,119],"edge":[33,47,86,129],"platforms":[34],"for":[35,46,85],"local,":[36],"secure":[37],"inference.":[38],"Binary":[39],"transformers":[40,59],"offer":[41],"a":[42,90,132,140],"compact,":[43],"low-complexity":[44],"solution":[45],"with":[49,99,149],"reduced":[50],"bandwidth":[51],"needs":[52],"acceptable":[54],"accuracy.":[55],"existing":[57],"binary":[58,70,82,93,147],"perform":[60],"inefficiently":[61],"on":[62,128],"current":[63],"hardware":[64],"due":[65],"the":[67,113,145],"lack":[68],"of":[69],"specific":[71],"optimizations.":[72],"To":[73],"address":[74],"this,":[75],"we":[76],"introduce":[77],"COBRA,":[78],"an":[79],"algorithm-architecture":[80],"co-optimized":[81],"Transformer":[83],"accelerator":[84],"computing.":[87],"COBRA":[88,116],"features":[89],"real":[91],"1-bit":[92],"multiplication":[94],"unit,":[95],"enabling":[96],"matrix":[97],"operations":[98],"-1,":[100],"0,":[101],"+1":[103],"values,":[104],"surpassing":[105],"ternary":[106],"methods.":[107],"With":[108],"further":[109],"hardware-friendly":[110],"optimizations":[111],"attention":[114],"block,":[115],"achieves":[117],"up":[118],"3,894.7":[120],"GOPS":[121],"throughput":[122,142],"448.7":[124],"GOPS/Watt":[125],"energy":[126,134],"efficiency":[127,135],"FPGAs,":[130],"delivering":[131],"311\u00d7":[133],"improvement":[136,143],"over":[137,144],"GPUs":[138],"3.5\u00d7":[141],"state-of-the-art":[146],"accelerator,":[148],"only":[150],"negligible":[151],"inference":[152],"accuracy":[153],"degradation.":[154]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-11-20T00:00:00"}